Evaluation of the effects of hemolysed cell type and number on biochemical test results during the in vitro hemolysis process

Hemolysis, the most frequent preanalytical error in clinical laboratories, is an interference which has very variable effects on biochemical test results. The aim of this study is to examine the effects of the hemolysed cell type and number as a reason for this variability using an experimental design. One unit of blood was obtained from 10 volunteers and RBC and PLT suspensions were prepared using quadruple blood bag system, WBC suspensions were prepared from the blood left in the primary bag from each subject. Cell suspensions were washed with serum physiological followed by serial dilutions to obtain seven different concentrations and then the hemolysis of cells was achieved by freeze-thaw cycle. Serum physiological and cell hemolysates were added (1:20) to the serum pool prepared from the serum samples of 3 volunteers to obtain the basal and seven different concentrations respectively. The 12 biochemical test parameters (LDH, potassium, AST, iron, CK, magnesium, total protein, total cholesterol, inorganic phosphorus, glucose, calcium ve BUN) and serum indices were measured by Beckman Coulter AU5800 autoanalyser. As cell lysis markers, free hemoglobin for RBC was measured by spectrophotometrically; MPO for WBC and β-TGB for PLT were measured by ELISA methods manually. The levels of biochemical test parameters were compared to basal sample using "Wilcoxon Signed Rank Test". Additionally, the relative change of each parameter according to the basal sample was calculated. The cell number which caused statistically significant difference was defined as the analytical threshold and the cell number which caused clinically significant bias was defined as the clinical threshold for each parameter. The RBC concentrations (number/µL) defined as the clinical threshold were 50000 for LDH, 100000 for AST, 200000 for potassium, 400000 for magnesium and inorganic phosphorous, 800000 for total protein, 162000 for CK and total cholesterol. The PLT concentration (number/µL) defined as the clinical threshold was 90000 for LDH and total protein while the analytical threshold was 12000 for LDH and 90000 for potassium. The effect of WBC lysis could not be evaluated since WBC suspensions were contaminated by other cell types. As a cell lysis marker, fHb was significantly correlated with RBC number (r=0.999, p=0,001), while MPO and β-TGB showed no significant correlation to WBC and PLT numbers respectively ( p=0,436 and p=0,277). The findings of this study point out that the hemolysed cell type and number contribute to the influence of hemolysis in a variable manner and thus may be a reason for the variability of the biochemical test results in hemolysed samples. Further studies that explain the mechanisms of hemolysis will provide important findings to solve this common interference in clinical laboratories.